<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Michael Solidum on Medium]]></title>
        <description><![CDATA[Stories by Michael Solidum on Medium]]></description>
        <link>https://medium.com/@michael.solidum?source=rss-39fdc6607b28------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*IgrxXxVNA9lXcWo_t3Sg5Q.png</url>
            <title>Stories by Michael Solidum on Medium</title>
            <link>https://medium.com/@michael.solidum?source=rss-39fdc6607b28------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 05:48:04 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@michael.solidum/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Your Local Tysons Corner Partner for Branded Apparel and Promo Products]]></title>
            <link>https://medium.com/@michael.solidum/your-local-tysons-corner-partner-for-branded-apparel-and-promo-products-c52c7162b016?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/c52c7162b016</guid>
            <category><![CDATA[fully-promoted]]></category>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Wed, 12 Aug 2026 14:20:06 GMT</pubDate>
            <atom:updated>2026-08-12T14:20:06.195Z</atom:updated>
            <content:encoded><![CDATA[<p>In the heart of Tysons Corner, a local business is dedicated to helping your brand stand out. <a href="https://fullypromoted.com/locations/tysonscorner-va">Fully Promoted Tysons Corner</a>, located at 8609 Westwood Center Dr. Suite 110 in Vienna, VA, is your go-to expert for custom apparel, promotional products, and signage. We are a proud part of the world’s largest branded apparel franchise, combining global resources with the personalized service of a local, community-focused team.</p><h3>Why Local Businesses Choose Fully Promoted Tysons Corner</h3><p>Whether you are a startup in Tysons or an established company in Vienna, our goal is to make your brand memorable. We understand that your brand is your promise to your customers, and we are here to help you deliver it with quality and consistency.</p><p>Simple and Fast Service: We make the ordering process straightforward and efficient. We are known for fast turnaround times and clear communication, ensuring your order is ready when you need it.</p><p>High-Quality Results: Quality matters. Our services include custom embroidery and screen printing on a wide range of apparel, ensuring your logo looks professional and lasts. We only use top-quality products and the best branding processes.</p><p>One-Stop Shop: From team uniforms and employee appreciation gifts to promotional items for events and trade shows, we have you covered. We can brand everything from apparel and bags to drinkware and pens.</p><blockquote>“If you need branded apparel for your company this is the place to go in Tyson’s Corner. Good prices and fast service.”<em> — VA (5-star review)</em></blockquote><h3>Let’s Get Your Brand Noticed</h3><p>Ready to elevate your brand? We invite you to visit our showroom, feel the fabrics, and see the quality for yourself. Our local experts are ready to listen to your needs and find the perfect branding solutions for your budget and timeline.</p><p>Visit us at 8609 Westwood Center Dr. Suite 110, Vienna, VA, 22182. We are open Monday through Friday from 10:00 a.m. to 4:00 p.m. and Friday from 9:00 a.m. to 3:00 p.m. Give us a call at 571–520–0655 to get a quote or start your next project today.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c52c7162b016" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[McLean Forrester: Your Strategic Partner for Practical AI and IT Modernization in 2026 and Beyond]]></title>
            <link>https://medium.com/@michael.solidum/mclean-forrester-your-strategic-partner-for-practical-ai-and-it-modernization-in-2026-and-beyond-bea8dbf5b7ba?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/bea8dbf5b7ba</guid>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Thu, 16 Jul 2026 15:55:11 GMT</pubDate>
            <atom:updated>2026-07-16T15:55:11.225Z</atom:updated>
            <content:encoded><![CDATA[<p>In a business landscape where artificial intelligence has moved from a futuristic concept to an operational necessity, organizations face a critical challenge. The era of AI experimentation is over, and the time for practical, results driven implementation has arrived. As we navigate 2026 and look toward the future, the question is no longer whether to adopt AI, but how to do so effectively, securely, and with a clear return on investment. This is where McLean Forrester emerges as an indispensable partner, offering a unique blend of deep technical expertise, strategic vision, and a people first philosophy that sets it apart in a crowded field of technology consultants.</p><h3>The McLean Forrester Difference: Where AI Gets Practical</h3><p>McLean Forrester, under the leadership of CEO Heather McLean, has built its reputation on a straightforward but powerful premise: making AI work the way it should. This means cutting through the hype and focusing on what truly matters for business leaders, delivering clarity, speed, and a tangible strategic edge. The company, a woman and veteran owned IT consulting firm based in the St. Louis area, brings over 25 years of technology experience to the table, including Heather McLean’s impressive 20 year career in the United States Air Force. This military background instills a mission first mindset, a focus on logistics and operational efficiency, and an unwavering commitment to the success of the client’s mission.</p><p>What truly distinguishes McLean Forrester is its philosophy. The firm does not start with technology; it starts with the business problem. As Heather McLean often emphasizes, the process always begins with a deep dive into the client’s unique challenges and operational bottlenecks. This client centered approach ensures that AI is not adopted for its own sake but as a targeted solution to a specific, well understood problem. It is a philosophy that resonates deeply in 2026, where businesses are demanding tangible results and questioning investments that do not deliver clear value.</p><h3>A Foundation Built on Three Pillars: Security, Simplicity, and Real Outcomes</h3><p>The entire McLean Forrester methodology is structured around three core pillars that ensure any AI or IT initiative is not just innovative, but also trusted, adopted, and successful.</p><p><strong>Security as the Foundation of Trust</strong></p><p>In an era of increasing data breaches and stringent regulations, security is non-negotiable. Many organizations hesitate to embrace AI due to fears about data privacy and compliance, a concern that remains a primary barrier to adoption. McLean Forrester tackles this head-on with an “Enterprise Secure AI” approach. They architect solutions that live within the client’s controlled environment, whether on-premise or in a private cloud. This is particularly critical for regulated sectors like healthcare, finance, and government. By prioritizing data integrity from the ground up and deploying solutions in secure, private cloud environments, McLean Forrester ensures that innovation does not come at the cost of compliance or security. This commitment is a cornerstone of their value proposition, allowing clients to move forward with confidence.</p><p><strong>The Critical Role of Simplicity and User Adoption</strong></p><p>A powerful AI tool is useless if employees find it too complex or intimidating to use. This is a lesson many organizations have learned the hard way. McLean Forrester prioritizes creating intuitive interfaces and seamless integrations that feel like a natural part of an employee’s workflow. The goal is to make work easier, not more complicated. By providing comprehensive training and empowering the workforce to become “agent architects” who can manage AI workflows, they drive adoption and ensure a return on investment. As Heather McLean succinctly puts it, if the solution is not simple, it will not be adopted, and if it is not adopted, it cannot possibly deliver value.</p><p><strong>An Obsession with Tangible, Measurable Outcomes</strong></p><p>Perhaps the most defining pillar of McLean Forrester’s approach is its unwavering focus on real outcomes. This is what separates them from a theoretical experiment or a vague promise of “digital transformation.” Success is defined in concrete terms from the very first day of any engagement. Is the goal a 40 percent reduction in the time it takes to generate monthly reports? A 20 percent improvement in marketing conversion rates? A significant decrease in customer complaint resolution times? By tying their work directly to these metrics, McLean Forrester ensures they are delivering tangible value and a clear strategic advantage. This focus on measurable value is increasingly critical as CFOs demand hard ROI and question flashy, unproven tech investments.</p><h3>Practical Solutions for Real World Challenges</h3><p>McLean Forrester’s expertise extends beyond<a href="https://mcleanforrester.com/services/ai-value-path/"> AI integration</a> to encompass a full spectrum of IT solutions designed to modernize and future-proof organizations. Their capabilities include application modernization, cloud migration, and IT and data strategy. They solve for the complete picture, addressing people, process, and technology to make business goals a reality.</p><p>A key offering that exemplifies their holistic approach is the AI Assisted 360 Degree IT Portfolio Analysis and Rationalization. This is a two phased methodology designed to strategically overhaul an organization’s entire IT portfolio. In the first phase, AI is used for a top down and bottom up analysis of the application portfolio, identifying security risks, cloud readiness, functionality, and system dependencies. The second phase leverages these insights to develop a comprehensive modernization roadmap, deciding whether to refactor, re-architect, rehost, rebuild, or replace each application. This service provides a complete overview, enabling organizations to identify opportunities for improvement that drive business value, mitigate risk, and lay the groundwork for future growth.</p><h3>A People First Culture That Delivers Results</h3><p>The effectiveness of <a href="https://mcleanforrester.com/">McLean Forrester</a> is rooted not just in its technology but in its culture. The company operates on a 36 hour work week and is fully virtual, decisions made to attract top talent and prevent burnout by prioritizing people over rigid corporate structures. This people first philosophy extends directly to client relationships. As Heather McLean states, they choose to prioritize their people and customers, and while they need to make a profit, it is not the driving force behind decision-making. This approach is validated by their clients. Testimonials from partners like the IT Director at Integrity Doctors LLC and the City Clerk of Cameron, MO, consistently praise the firm’s excellent communication, transparency, and the game-changing value they deliver.</p><h3>Looking Ahead: Your Partner for 2026 and Beyond</h3><p>As we move further into 2026 and beyond, the business world will only become more complex and technologically driven. The convergence of AI with other emerging technologies will create new opportunities and challenges. For leaders who feel the pressure to act but are wary of the risks, McLean Forrester offers a reliable and proven path forward. They provide the clarity to see through the hype, the speed to capture opportunities, and the security to operate with confidence.</p><p>By focusing on where AI gets practical, Heather McLean and her team are not just helping clients transition into the future; they are ensuring that when the technology is finally doing its job, the business is free to do its best work. Whether you are a local municipality struggling with outdated operations and have saved 40% on your IT budget through modernization, or a financial services firm looking to automate document processing, McLean Forrester builds the practical, secure, and effective solutions you need to thrive. In a landscape full of empty promises, McLean Forrester stands as a beacon of practical, people centered, and results driven innovation.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bea8dbf5b7ba" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The NDAA Technical Debt Challenge: Why the Status Quo Is Failing and How to Fix It]]></title>
            <link>https://medium.com/@michael.solidum/the-ndaa-technical-debt-challenge-why-the-status-quo-is-failing-and-how-to-fix-it-8959fd68e04c?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/8959fd68e04c</guid>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Wed, 15 Jul 2026 15:41:16 GMT</pubDate>
            <atom:updated>2026-07-15T15:41:16.051Z</atom:updated>
            <content:encoded><![CDATA[<p>The recently published National Defense Authorization Act, or NDAA, has placed a sharp focus on a problem that has quietly plagued the Department of Defense for decades: technical debt. While the NDAA lays out clear requirements for reducing legacy system risk, modernizing software portfolios, and improving cybersecurity hygiene, the reality on the ground is far more complicated.</p><p>Across the DoD enterprise, there is a growing recognition that technical debt is not just an IT nuisance. It is a strategic vulnerability. It slows down mission execution, creates security gaps, and wastes billions in maintenance costs. But while the problem is widely acknowledged, the path forward remains murky. Why? Because the burden of solving it is falling on groups that are not equipped to handle it.</p><p>Let us look at how the current landscape is structured and why it is failing.</p><p><strong>The Operators: A Misplaced Trust</strong></p><p>For the operators, the warfighters and mission leads who depend on reliable systems every day, there is a quiet assumption that the CIO team has this under control. They believe that somewhere in the background, a dedicated group of technologists is systematically identifying outdated code, decommissioning obsolete applications, and streamlining the portfolio.</p><p>But that belief is largely misplaced. CIO teams are often understaffed, overstretched, and buried in day-to-day operational firefighting. They do not have the bandwidth to conduct deep application rationalization across thousands of systems. More importantly, they often lack the authority to make decisions about which programs stay and which go. That authority rests with the program owners, and that is where the real friction begins.</p><p><strong>The Cyber Professionals: Flying Blind</strong></p><p>For the cyber professionals, the challenge is even more acute. They are tasked with protecting systems they do not fully understand. They do not have access to the detailed system information owned by individual programs. They cannot see the codebase, the dependency maps, or the data flows that would allow them to assess true risk.</p><p>This creates a dangerous paradox. The people responsible for defending the network are often the last to know what is actually running on it. They are forced to rely on high level reports and incomplete inventories, which means their vulnerability assessments are only as good as the data they receive. And that data is rarely complete.</p><p><strong>The Portfolio Managers: Stuck in Self Reporting</strong></p><p>Then we have the portfolio managers. These are the individuals tasked with overseeing multi billion dollar portfolios of IT systems. They are accountable for cost, performance, and security. But they are operating with one hand tied behind their backs.</p><p>Why? Because they depend on the programs to self report their status. They ask for data on system health, modernization progress, and technical debt levels. But the programs are incentivized to paint a rosy picture. Admitting to high technical debt could threaten funding or future budgets. So the portfolio managers receive polished reports that often mask the true scale of the problem.</p><p>This is not a failure of people. It is a failure of process. The entire ecosystem relies on trust and self reporting, which is not a sustainable model for something as critical as technical debt reduction.</p><p><strong>A Huge Challenge, but Not an Impossible One</strong></p><p>When you step back and look at the big picture, it is easy to see why the NDAA requirements feel overwhelming. You have operators who assume someone else is handling it. You have cyber professionals who cannot get the data they need. You have portfolio managers who are forced to rely on incomplete or biased information.</p><p>And in the middle of all this, you have legacy systems that were built decades ago, running on outdated languages, with no clear owner, no modern documentation, and no easy path to retirement.</p><p>This is the technical debt trap. And it is costing the DoD far more than just money. It is costing agility, security, and mission readiness.</p><p><strong>A Better Way Forward</strong></p><p>At McLean Forrester, we have spent years studying this exact problem. We have worked with defense agencies, federal civilians, and system integrators to understand where the bottlenecks are and how to break through them. And we have built a solution that directly addresses the root causes of the technical debt crisis.</p><p>Our Application Rationalization 360° solution, or APP RAT 360°, is currently posted on Tradewinds, the DoD’s marketplace for innovative commercial solutions. And Tradewinds has already validated what we have known all along. They have called our solution unique, innovative, and officially “Awardable.” That is not just marketing language. That is a formal recognition from the very platform that the DoD uses to vet new technologies.</p><p><strong>What Makes APP RAT 360° Different?</strong></p><p>First, we do not rely on self reporting. Our platform ingests actual system data, not spreadsheets. We look at code repositories, configuration files, network traffic, and usage patterns. We generate a real time, fact based picture of every application in your portfolio.</p><p>Second, we give cyber professionals the visibility they have been missing. For the first time, they can see exactly what is running, what it connects to, and what the true risk profile looks like. No more blind spots. No more guessing.</p><p>Third, we empower portfolio managers with actionable insights, not just reports. We help them identify which applications are safe to retire, which ones need to be modernized, and which ones are actually mission critical. And we do this in a fraction of the time it would take using traditional methods.</p><p><strong>Speed, Cost, and Trust</strong></p><p>You might be wondering about cost. The reality is that our solution costs far less than the resources you have likely already put toward this task. We are not asking you to hire new teams or launch multi year studies. We are offering a proven, off the shelf capability that delivers results in days, not years.</p><p>Time is another factor. Most organizations assume that application rationalization takes months or even years. We have seen the opposite. In many cases, we can deliver initial findings within hours and comprehensive assessments within days. That is not a promise. That is what we have already done for other clients.</p><p>And trust? Gartner, one of the most respected research firms in the world, has named our tool suite as a Technical Debt Reduction Magic Quadrant Leader. That is not an endorsement we take lightly. It is a third party validation that our approach is not just different, but better.</p><p><strong>The NDAA Section 1503 Crosswalk</strong></p><p>We understand that compliance is a major driver for many of our clients. That is why we have created a detailed crosswalk that maps every single NDAA Section 1503 requirement directly to our <a href="https://mcleanforrester.com/app-rat-rapid-it-portfolio-analysis-rationalization/">APP RAT 360°</a> capabilities. You can see exactly how we address each mandate, from system inventory to risk scoring to modernization planning.</p><p>That crosswalk is available for download in the first comment of this post. We encourage you to take a look and see for yourself how our solution aligns with your specific requirements.</p><p><strong>Let Us Talk</strong></p><p>The challenges we have outlined here are real. They are not going away on their own. But they are solvable. And we are here to help.</p><p>Our team at <a href="https://mcleanforrester.com/">McLean Forrester</a> has the experience, the technology, and the track record to guide you through this process. We have helped agencies reduce their technical debt by more than 40 percent in under six months. We have helped cyber teams close critical vulnerabilities that had been open for years. And we have helped portfolio managers finally get the accurate, trustworthy data they need to make informed decisions.</p><p>Chats are free. There is no obligation, no pressure, and no sales pitch. Just an honest conversation about your challenges and how we might be able to help.</p><p>You can reach me directly at Larry.McLean@mcleanforrester.com. Or visit our website to learn more about our solutions and case studies.</p><p><strong>Your Next Step</strong></p><p>Technical debt is not going to fix itself. The NDAA requirements are not going to get easier. And the status quo is not going to suddenly start working.</p><p>But with the right partner, the right tools, and the right approach, you can turn this challenge into an opportunity. You can reduce risk, save money, and improve mission outcomes. And you can do it faster than you ever thought possible.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8959fd68e04c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The AI Governance Trap: Why One-Size-Fits-All Supervision Is Setting Your Enterprise Up for Failure]]></title>
            <link>https://medium.com/@michael.solidum/the-ai-governance-trap-why-one-size-fits-all-supervision-is-setting-your-enterprise-up-for-failure-cd993d08662d?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/cd993d08662d</guid>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Tue, 14 Jul 2026 15:38:13 GMT</pubDate>
            <atom:updated>2026-07-14T15:38:13.641Z</atom:updated>
            <content:encoded><![CDATA[<p>Let’s start with a simple question: Is your AI agent technically supervised?</p><p>If you answered yes, congratulations. You have checked the box.</p><p>But here is the uncomfortable truth that Gartner recently dropped on the enterprise world. Technically supervised is not the same thing as appropriately governed. And if your organization is applying the same governance model to every AI agent, you are likely creating two very different and equally predictable problems .</p><p>Simple agents get buried in controls they do not need. Powerful autonomous agents get nowhere near the oversight they actually do need.</p><p>It sounds ridiculous when you think about it in human terms. Imagine taking a group of kindergartners on a field trip. One is playing in the sandbox. One is writing a note. One is driving the school bus. And one is launching a rocket. But do not worry, the chaperone checked the box. Everyone is supervised.</p><p>That is absurd. Yet it is not far from how some organizations are approaching AI agents today. One policy. One approval process. One set of controls. Regardless of what the agent can access or what it can actually do.</p><p>The stakes here are real. Gartner predicts that by 2027, 40 percent of enterprises will demote or decommission autonomous AI agents because governance gaps are discovered only after production incidents . That is not a warning about the future. That is a forecast about what is already happening.</p><p>The question is not simply whether you govern your AI agents. The question is whether you are governing each agent according to what it is trusted and able to do.</p><h3>The Binary Trap</h3><p>The root cause of this failure is that enterprises are treating AI agent governance as binary . Agents are either locked down completely or fully trusted. There is no middle ground, no nuance, no recognition that different agents have different capabilities and different risk profiles.</p><p>This binary thinking leads to two common failure modes .</p><p>First, over-restriction of simple agents. A summarization bot gets the same approval workflow as a system configuration agent. It slows delivery, frustrates users, and drives shadow development as teams create ungoverned workarounds to get their jobs done.</p><p>Second, under-restriction of powerful agents. A fully autonomous agent gets the same minimal oversight as a read-only assistant. This increases operational, security, and compliance risk. And because the governance gaps are only discovered after something goes wrong, the damage is already done.</p><p>As Shiva Varma, Senior Director Analyst at Gartner, put it, agents operate at different autonomy levels and across different trust boundaries . When the same controls are applied indiscriminately, you get the worst of both worlds.</p><h3>Proportional Governance: The Gartner Framework</h3><p>Gartner recommends a proportional governance approach that classifies AI agents across four distinct autonomy levels . Each level represents a different trust boundary and requires corresponding governance requirements.</p><h3>Level 1: Observe</h3><p>At this level, agents are limited to read-only access to defined data sources. Their outputs are visible only to the requesting user. Common use cases include document summarization, data or knowledge retrieval, and code explanation .</p><p>Governance at this level is straightforward. It focuses on baseline controls like scoped data access, user authentication, usage logging, and basic functional and security testing. Because the risk is limited primarily to data exposure and output accuracy, controls should remain lightweight and targeted .</p><h3>Level 2: Advise</h3><p>Advise agents generate recommendations, drafts, or proposed actions. Humans review all outputs and execute actions manually. These agents retain read-only access with no write access to any system. They are commonly used for email drafting, report or code generation, and decision support .</p><p>Here is the catch. Even though humans execute the final actions, advisory agents can anchor judgment. When inaccurate outputs are trusted due to automation bias, downstream risk emerges .</p><p>Governance for advise agents must extend beyond Level 1 controls. It should include accuracy and hallucination testing, domain-specific quality evaluations, and user training on appropriate reliance levels .</p><h3>Level 3: Act with Approval</h3><p>At Level 3, agents can execute actions like writing data, sending communications, or modifying configurations. But they can only do so after explicit human approval for every action .</p><p>This is where governance gets tricky. Human review is effective only if it remains a meaningful control. Without strong security testing, clear approval workflows with audit trails, and agent-specific incident response procedures, approvals degrade under time pressure or approval fatigue . You end up with a false sense of safety while actually expanding your attack surface.</p><h3>Level 4: Act Autonomously</h3><p>At the highest autonomy level, agents execute actions independently within defined guardrails. Humans review exceptions, audit logs, and aggregated outcomes rather than individual decisions .</p><p>When agents operate autonomously, actions are executed at a scale and speed that can outpace human oversight. Because accountability for outcomes remains with the organization, this level requires the most rigorous governance. That includes continuous monitoring, enforced guardrails, rapid rollback mechanisms, circuit breakers that halt agent operation on threshold violations, and clear ownership for agent behavior .</p><h3>The Governance Gap Is Wider Than You Think</h3><p>The Gartner warning is not happening in isolation. Recent research from BSI found that less than a quarter of organizations reported having an AI governance program in place . Even among large enterprises, that number only rose to about a third.</p><p>EY research tells a similar story. While nearly three-quarters of executives say their organizations have integrated AI across most or all initiatives, only a third have the proper protocols to adhere to responsible AI frameworks . The C-suite is confident, but confidence is not the same thing as preparation.</p><p>The disconnect gets worse. CEOs are actually more concerned about AI risks than other C-suite leaders. But that concern is not translating into action. Governance is lagging behind innovation, and half of executives admit it is challenging to develop governance frameworks for current AI technologies, let alone the next generation .</p><p>This is not a niche concern. It is a widespread gap that spans industries, geographies, and organization sizes.</p><h3>What Proportional Governance Actually Looks Like</h3><p>Proportional governance means scaling controls with the agent’s autonomy, the sensitivity of its access, the potential impact of its actions, and the difficulty of reversing a mistake .</p><p>A document summarization tool does not need the same governance as a system configuration agent. An agent that drafts emails does not need the same oversight as an agent that sends customer communications. An agent that recommends actions does not need the same controls as an agent that executes them.</p><p>The governance should fit the risk. Not the other way around.</p><p>This is where frameworks like the IETF AI Governance and Accountability Protocol (AIGA) come into play . They provide a structured way to classify agents into risk tiers and apply proportional oversight. AIGA distinguishes between Tier 0 minimal agents like chatbots, Tier 1 low-risk agents like code assistants, Tier 2 medium-risk agents like trading bots, Tier 3 high-risk agents operating critical infrastructure, and Tier 4 restricted agents with the highest potential for harm.</p><p>Each tier has different technical requirements for trusted computing base, approval latency, and check-in frequency . The governance is not one-size-fits-all. It scales with the risk.</p><h3>Getting Practical: Governance in Action</h3><p>So what does this look like in practice? Here are some concrete steps for moving from binary governance to proportional governance.</p><p>First, inventory your agents. You cannot govern what you do not know exists. According to MIT Technology Review, every agent should run as a non-human principal with the same discipline applied to employees . Can you list every agent in your environment and exactly what each is allowed to do?</p><p>Second, classify agents by autonomy. Not all agents are the same. Some are read-only. Some make recommendations. Some execute actions with approval. Some operate autonomously. Each category needs different controls.</p><p>Third, enforce controls at the boundaries. The controls that matter are the ones at the identity boundary, the tool boundary, the data boundary, and the output boundary . This means making agents authenticate properly, pinning tool versions and requiring approvals for new tools, treating external content as hostile until proven otherwise, and validating outputs before they act.</p><p>Fourth, build for reversibility. When autonomous agents act at scale and speed, mistakes will happen. The question is whether you can undo them. Rapid rollback mechanisms, circuit breakers, and clearly defined accountability are not optional for high-autonomy agents .</p><p>Fifth, monitor continuously. Governance is not a one-time exercise. As MIT Technology Review notes, enterprises must have a living catalog and unified logs that show which agents exist, what scopes and tools each is allowed, and every high-impact action with who approved it and when .</p><h3>Why This Matters for Your Business</h3><p>This is not a compliance exercise. This is a business imperative.</p><p>The organizations that get governance right will scale AI agents safely and capture the productivity benefits. The ones that get it wrong will find themselves demoting or decommissioning their autonomous agents after incidents that could have been prevented .</p><p>Governance failures are expensive. They expose organizations to operational risk, security breaches, and regulatory penalties. But they also create a hidden cost: lost opportunity. When simple agents are over-restricted, innovation slows. When powerful agents are under-restricted, the trust deficit grows. Either way, the organization loses.</p><p>The good news is that governance is not a mystery. Frameworks exist. Standards are emerging. And the core principle is straightforward: governance should scale with risk.</p><p>Technically supervised may be enough for the sandbox. But it is not enough for the launchpad.</p><h3>Where to Start</h3><p>If you are reading this and realizing your organization is in the binary trap, do not panic. You are not alone. Most organizations are still figuring this out.</p><p>Start with the basics. Inventory your agents. Classify them by autonomy level. Apply controls that fit the risk. Build for monitoring and reversibility. And keep asking the hard question: Are we governing this agent according to what it is trusted and able to do?</p><p>The companies that answer that question honestly will be the ones that succeed with AI agents. The ones that keep checking the box will be the ones Gartner is talking about.</p><p>The choice is yours.</p><p><em>For more insights on AI governance and implementing practical solutions that balance innovation with risk management, visit </em><a href="https://mcleanforrester.com/"><em>McLean Forrester</em></a><em>. Their work on </em><a href="https://mcleanforrester.com/"><em>application modernization</em></a><em> shows how organizations can build the foundation needed for AI to deliver on its promise.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cd993d08662d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Org Side of AI: Why Copilot Cowork Succeeds or Fails Before Anyone Clicks “Accept”]]></title>
            <link>https://medium.com/@michael.solidum/the-org-side-of-ai-why-copilot-cowork-succeeds-or-fails-before-anyone-clicks-accept-878f3bacc5aa?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/878f3bacc5aa</guid>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Mon, 13 Jul 2026 15:51:13 GMT</pubDate>
            <atom:updated>2026-07-13T15:51:13.748Z</atom:updated>
            <content:encoded><![CDATA[<p>Microsoft just launched Copilot Cowork, an agentic layer that executes multi-step workflows across your entire M365 suite, and your org just bought licenses. The sales pitch is seductive. An AI agent that moves between Outlook, Teams, Excel, and SharePoint, automating the workflows that eat up hours of your employees’ days. The deployment timeline is measured in weeks. The governance framework? That gets built in hindsight, if at all.</p><p><a href="https://www.blogger.com/blog/post/edit/36927854850194242/6426809328413242530#">Heather McLean</a>, CEO of McLean Forrester, has seen this pattern play out too many times to count. She has observed it across organizations of every size and industry. The tools deploy fast. The organizational infrastructure to support them? That gets built in hindsight, if at all. And that is where the real value gets lost.</p><p>According to Heather, the problem is not the technology. She has written extensively about the biggest misconception about AI, and her message is clear: “It’s not about the technology.” CIOs and CTOs who focus on the tools often miss the bigger picture. The real starting point is defining the business outcome you want. Only then do you identify the right projects. Only then does the tech matter .</p><p>Copilot Cowork is fundamentally different from traditional software. It is an agentic system that executes multi-step workflows autonomously. It does not just sit in a tool waiting for commands. It acts. It reasons. It makes decisions within defined parameters. This changes the calculus for workflows, governance, and change management in ways that most organizations are not prepared for.</p><p>The gap Heather sees every day is between the speed of technology deployment and the pace of organizational readiness. The tools get deployed in weeks. The governance, the accountability structures, the retraining of your workforce on how to work alongside an AI agent that executes complex workflows, that takes months. Or it never happens at all.</p><p>The data supports her observation. According to McKinsey, 70% of business transformations fail. For AI projects specifically, the failure rate is even higher. Many organizations find themselves stuck in a cycle of pilot projects that never scale, or worse, they risk creating a fragmented and unmanaged ecosystem of AI tools . Heather points to these numbers constantly when she talks to clients. She tells them that the pattern is clear. Organizations that succeed with AI are the ones that invest as much in the organizational side as they do in the technology side.</p><h3>Rethinking Workflows: Automating the Broken Process Is Not a Win</h3><p>Copilot Cowork excels at multi-step workflows. But here is the question Heather says gets skipped every time. Are those workflows worth automating in their current form? She has seen organizations rush to automate processes that were broken to begin with. They end up with an AI agent that executes flawed workflows at machine speed. That is not value creation. That is compounding inefficiency.</p><p>The right approach starts with workflow re-engineering. Before you let Copilot Cowork loose on your M365 environment, you need to map the workflows it will execute. Identify the bottlenecks. Question the handoffs. Eliminate the steps that exist only because “that is how we have always done it.” This is not about technology. This is about operational clarity.</p><p>Heather comes from a logistics background, 25 years in technology and logistics, with a 20-year military career. She views automation as a logistics problem, the efficient movement of information and orchestration of complex workflows . If you do not fix the underlying flow, you are just automating chaos. She has stated that this logistics mindset is what sets her approach apart from other consultants who focus purely on the technology.</p><p>At <a href="https://www.blogger.com/blog/post/edit/36927854850194242/6426809328413242530#">McLean Forrester</a>, the process always starts with the business goal. They define outcomes, identify pain points and opportunities, prioritize high-value project candidates, and then prototype, validate, and scale . Heather emphasizes that they always start by asking questions about the business, not about AI. Where are the inefficiencies? What processes are consuming too many valuable human hours? Where is data being collected but not utilized for decision making? Once they thoroughly understand the problem, they can then evaluate if AI is the right tool for the job .</p><h3>Governance: Who Has Accountability When the Agent Makes a Mistake?</h3><p>This is the question that keeps CIOs up at night. Copilot Cowork executes workflows. It takes actions. It generates outcomes. When something goes wrong, who is accountable? The employee who initiated the workflow? The IT team that deployed the agent? The vendor that built the model?</p><p>Heather has seen organizations struggle with this question repeatedly. She says the answer needs to be clear before the tool goes live. Governance for agentic AI requires a different framework than traditional software governance. You need to define decision rights, escalation paths, audit trails, and accountability structures.</p><p>In her work, Heather treats security and governance as non-negotiable. She has stated firmly that “implementing an AI solution without a robust security framework is not an implementation at all; it is an unacceptable risk” . Her team prioritizes data integrity from the ground up, deploying solutions in secure, private cloud environments, employing strict data anonymization practices, and carefully vetting technology partners. For a solution to be considered practical, it must first be considered safe .</p><p>She recommends defining these elements clearly:</p><ul><li>Decision rights: What decisions can the agent make autonomously? What requires human review?</li><li>Escalation paths: When does the agent flag a decision for human intervention?</li><li>Audit trails: How do you trace every action back to a clear provenance?</li><li>Accountability: Who owns the outcome of each workflow?</li></ul><p>Heather emphasizes that this is not a one-size-fits-all solution. Your governance framework needs to reflect your risk tolerance, your regulatory environment, and your organizational culture.</p><h3>Change Management: Bringing Your People Along</h3><p>Heather has a message that she repeats constantly. The biggest roadblock to AI success is not technical. It is human. Employees who find the technology intimidating or complex simply will not adopt it. And if they do not adopt it, you will not see a return on your investment.</p><p>Copilot Cowork represents a significant shift in how people work. Employees are not just using a new tool. They are collaborating with an AI agent that executes complex workflows. This requires a mindset shift. People need to trust the agent. They need to understand its capabilities and its limitations. They need to know when to defer to it and when to override it.</p><p>The old change management playbook does not work for AI. Heather has observed that traditional approaches focus on upfront engagement, communications, training, and managing stakeholder concerns. But for AI, the real work begins after go-live, when people try to integrate the tools into their daily work and often hesitate, struggle, or revert.</p><p>She emphasizes the critical role of simplicity in driving adoption. “The goal of AI should be to make work easier, not more complicated,” she has said. Her team focuses on creating intuitive interfaces and seamless integrations that feel like a natural part of an employee’s workflow. They provide training that empowers people to use the new tools effectively. Heather is clear about the consequence of ignoring this: “If the solution is not simple, it will not be adopted, and if it is not adopted, it cannot possibly deliver a return on investment” .</p><p>Heather emphasizes that change does not happen in a day. It happens through consistent reinforcement, clear incentives, and a culture that rewards experimentation. She has told me that this is where most organizations fall short. They invest in the initial training but fail to sustain the momentum.</p><h3>The Bottom Line: Leadership Beats Technology</h3><p>Heather has spent 20 years as an Air Force officer and another 25 years in technology. She has learned one lesson that applies to every project, leadership beats technology. Every single time.</p><p>Copilot Cowork is a powerful tool. It can automate complex workflows, free your employees from tedious tasks, and deliver measurable operational improvements. The Forrester Total Economic Impact study on Microsoft AI solutions found projected ROI ranging from 167% to 457%, with benefits including low-value task automation, supply chain agility, and reduced defects. Those numbers are compelling.</p><p>But the technology does not deliver value on its own. Your people do. Your governance framework does. Your commitment to re-engineering workflows and bringing your organization along does.</p><p>If your org is in the middle of an AI rollout and the organizational side is not keeping pace, you are not alone. But Heather would tell you that you need to act. The tools deploy in weeks. The organizational infrastructure takes time. The time to start is now.</p><p>At McLean Forrester, Heather and her team help clients bridge this gap every day. They start with the business problem, not the technology. They focus on security, simplicity, and real outcomes. They build governance frameworks that protect your organization while enabling innovation. And they design change management that actually works.</p><h3>Your Copilot Cowork licenses are purchased. Now the real work begins.</h3><p>Heather has seen the pattern across hundreds of organizations. The ones that succeed treat AI adoption as an organizational transformation, not a technology deployment. The ones that fail treat it as an IT project. The difference is not the tool. The difference is the commitment to rethinking workflows, building governance, and bringing people along.</p><p>The tools deploy in weeks. The organizational infrastructure to support them gets built in hindsight, if at all. Heather has made it her mission to change that pattern. She works with organizations to build the organizational infrastructure first, so that when the technology arrives, it actually delivers value.</p><p>If you are in the middle of an AI rollout and the organizational side is not keeping pace, Heather would tell you to pause. Take a step back. Build the governance. Re-engineer the workflows. Bring your people along. The technology will still be there. But your chance of success will be much higher.</p><p><em>You can learn more about the </em><a href="https://mcleanforrester.com/"><em>AI Value Path framework</em></a><em> or explore how </em><a href="https://mcleanforrester.com/"><em>Enterprise Secure AI</em></a><em> deployments can protect your organization while driving innovation.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=878f3bacc5aa" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The AI Playbook Has Changed Three Times in Four Years. Are You Still Using the Old One?]]></title>
            <link>https://medium.com/@michael.solidum/the-ai-playbook-has-changed-three-times-in-four-years-are-you-still-using-the-old-one-a5bdcc1235b2?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/a5bdcc1235b2</guid>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Mon, 06 Jul 2026 15:40:14 GMT</pubDate>
            <atom:updated>2026-07-06T15:40:14.107Z</atom:updated>
            <content:encoded><![CDATA[<p>The way we work with artificial intelligence has been completely rewritten three times in just four years. Not tweaked. Not updated. Rewritten from the ground up. What started as a simple exercise in writing better prompts has evolved into designing autonomous systems that find their own work, execute it, validate the results, and decide what to do next, all without a human sitting in the middle of every action. This is not a subtle shift in technique. This is a fundamental reorganization of how humans and machines collaborate. And if your organization is still focused on prompt engineering while the leaders in your industry are designing loops, you are already falling behind.</p><p>The playbook for working with AI has undergone three distinct transformations in that short span. Prompt engineering dominated the early days, when the primary challenge was figuring out how to phrase questions to get useful answers from large language models. Then came context engineering, which recognized that the quality of AI output depends less on clever prompting and more on providing the right information, the right data, and the right constraints. Now we are witnessing the rise of loop engineering, a paradigm that fundamentally changes the question we ask when working with AI.</p><p>Prompt engineering asked how to write the best instruction. Context engineering asked what information to give the model. Loop engineering asks a different question entirely: what system should we build so the AI finds the work, does it, evaluates the result, and determines its own next step, without a human in the middle of every action?</p><p>Peter Steinberger, the developer behind OpenClaw, articulated this shift when he said you should not be prompting coding agents anymore. You should be designing loops that prompt your agents. This is not a subtle distinction. It represents a complete reversal of how we think about human AI collaboration. Boris Cherny, who leads Claude Code at Anthropic, put it even more directly. He said loops are as big as the step from source code to agents was. Loops are the step from agents to the next thing. When the creator of one of the most widely used AI coding tools says he no longer writes prompts for Claude, that his loops prompt the system and decide what to do next, it is worth paying attention.</p><p>The evolution from prompt engineering to loop engineering follows a clear pattern. Every time the underlying technology takes a step forward, the way humans work with it has to take a step too. Prompt engineering, context engineering, and loop engineering each arrived because the previous model of working with AI had a ceiling. Prompt engineering could only do so much when the model lacked the right information. Context engineering solved that problem but could not prevent agents from making the same mistakes repeatedly or acting without oversight. Harness engineering added structure and observability, but it still required human triggering and judgment at each step. Loop engineering moves beyond all of these by building systems that run continuously, discovering work, assigning tasks, validating results, and deciding what to do next without requiring a person to press enter every time.</p><p>The core components of a loop are surprisingly straightforward, though the implications are profound. A complete loop requires five functional modules plus a memory layer. The first is automation, the heartbeat that keeps the loop running on a schedule or in response to events. The second is worktrees, which isolate parallel agents so they do not overwrite each other’s work. The third is skills, which encode project knowledge so agents do not have to rediscover your conventions and constraints with every run. The fourth is plugins and connectors, which give the loop access to real tools like issue trackers, databases, and communication platforms. The fifth is sub agents, which separate the work of generating from the work of verifying. A sixth element, memory, stores what has been completed and what comes next in a file that persists beyond any single conversation.</p><p>The sub agent pattern is particularly important. When the same model that writes code also reviews its own work, it tends to be too generous with itself. A separate agent with different instructions, sometimes using a different model entirely, can catch problems the first agent missed or chose to ignore. This is the structure that makes loops trustworthy enough to run while you are not watching. The verification agent is the reason you can walk away and let the system work.</p><p>Google’s Gemini Co Scientist, published this week in Nature, provides a powerful example of loop engineering applied outside software development. The system generates hypotheses, designs experiments, analyzes data, and refines its own research questions. The loop runs continuously, with specialized agents handling generation, reflection, ranking, and evolution. The system uses a tournament mechanism where hypotheses compete, are critiqued, and evolve over multiple iterations. In testing, Co Scientist successfully proposed new drug repurposing candidates for acute myeloid leukemia, identified novel therapeutic targets for liver fibrosis, and independently arrived at a mechanism for antimicrobial resistance that matched unpublished experimental findings.</p><p>What makes Co Scientist significant is not just the results but the architecture. The system is built on a multi agent framework where each agent has a specific role. The generation agent proposes initial hypotheses grounded in scientific literature. The reflection agent acts as a virtual peer reviewer, evaluating hypotheses for correctness and novelty. The ranking agent orchestrates an idea tournament, using pairwise comparisons and simulated scientific debates to prioritize the most promising paths. The evolution agent continuously refines and combines top ranked hypotheses. The meta review agent synthesizes insights and generates the final research proposal. A supervisor agent acts as an adaptive planner, breaking down high level goals into executable steps and coordinating agents to run in parallel.</p><p>The result is a system that does not just retrieve information but actively participates in scientific discovery. It generates novel, testable hypotheses that experts rated as more original and impactful than those from other leading models. In one case, the system identified a drug candidate that blocked 91 percent of a scarring linked response in liver fibrosis lab tests. In another, it proposed a combination therapy for AML that showed significant synergistic effects in vitro. These are not theoretical outputs. They are experimentally validated discoveries that could lead to new treatments.</p><p>The implications of loop engineering extend far beyond coding or scientific research. Any organization that uses AI to solve complex, recurring problems can benefit from this pattern. The key is to recognize that the goal is not to build a smarter chatbot but to design a system that autonomously pursues objectives. The human role shifts from operator to architect and supervisor. Your job is no longer to prompt the AI at each step. Your job is to design the loop, define the goals, set the boundaries, and validate the outcomes.</p><p>This shift comes with new challenges. Validation remains your responsibility. Even with verification agents, the final decision to accept or reject the loop’s output rests with you. Understanding debt grows as agents produce more output than you can personally read. The more smoothly the loop runs, the less you may understand the system it is building. Cognitive surrender is a real risk. When the loop works consistently, it is tempting to stop thinking critically about its outputs. The loop is a tool, not a replacement for judgment.</p><p>The costs are also real. Loops burn tokens continuously. The goal is to keep them running indefinitely, which means there is no ceiling on what you could spend. However, when the benefits outweigh the costs, as they clearly do for the organizations deploying Co Scientist or running thousands of agents overnight, the investment is justified.</p><p>The evolution from prompt engineering to loop engineering represents a fundamental shift in how we work with AI. The technology is moving from a tool you operate to a system that operates itself, with you as the architect and supervisor. Have you started to experiment with loops yet? The companies that embrace this shift early will define how AI transforms work in the coming years. Those that wait may find themselves playing catch up in a world where the question is no longer what to prompt but what to build.</p><p>To learn more about how organizations are structuring their AI initiatives for real business outcomes, explore the AI Value Path framework and the practical approaches to enterprise AI adoption available at <a href="https://mcleanforrester.com/">mcleanforrester.com</a>. For organizations ready to move beyond pilots and into production, the AI Value Path offers a disciplined, evidence-based approach to turning AI from an experiment into a competitive advantage.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a5bdcc1235b2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[From Chaos to Clarity: A Structured Approach to AI That Actually Delivers]]></title>
            <link>https://medium.com/@michael.solidum/from-chaos-to-clarity-a-structured-approach-to-ai-that-actually-delivers-c16519d6dff6?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/c16519d6dff6</guid>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Thu, 02 Jul 2026 15:32:42 GMT</pubDate>
            <atom:updated>2026-07-02T15:32:42.841Z</atom:updated>
            <content:encoded><![CDATA[<p>A leadership team had been wrestling with artificial intelligence for over a year. They had launched three separate pilots. They had hired a data science consultant. They had attended the conferences and read the analyst reports. But when I asked what had actually moved the needle on their business performance, the silence in the room spoke volumes. They were burning fuel in every direction with no clear vector. All thrust and no vector. And the answer was not to buy more AI tools.</p><p>The Real Gap in AI Adoption</p><p>This scenario is becoming the norm. Organizations feel the pressure to act. Competitors are deploying AI. Boards are demanding a strategy. Investors are asking tough questions about AI readiness. So companies start moving, often without a clear destination, and they end up with a collection of disconnected experiments that never scale to production. The failure rates are sobering. Research suggests that the majority of generative AI projects never make it out of the pilot phase. That is not because the technology does not work. It is because the conditions for success were never put in place.</p><p>Most organizations are trying to layer AI on top of foundations that simply cannot support it. Their data is scattered and inconsistent. Their application portfolios have grown organically over decades, creating technical debt that slows everything down. Their teams are spending too much time maintaining legacy systems and not enough time on innovation. When you point an AI model at that kind of environment, you do not get transformation. You get a faster way to surface existing problems.</p><p>The Discipline of Starting with “Not Yet”</p><p>The rule Heather McLean brings to every engagement is simple. The team does not move into build mode until the vision genuinely becomes theirs. When they can see what the client sees, they attack the solution with real focus, driving outcomes in customers’ hands rather than chasing outputs. And sometimes that means telling a client “not yet” before they ever write a line of code.</p><p>This is exactly what the <a href="https://mcleanforrester.com/ai-value-path">AI Value Path</a> from McLean Forrester was designed to do. It is a structured, low-risk path from AI exploration to full-scale deployment, organized across three phases with measurable outcomes at every step.</p><p>Phase One: Executive Alignment and Opportunity Prioritization</p><p>The first phase is not about technology. It is about people and priorities. For two weeks, the team conducts a forensic dive into your business. They sit with your executive team, but also with the people on the front lines who deal with operational headaches every day. They talk to your operations manager who understands where the supply chain friction points are. They sit with your customer service director who knows exactly where customers get frustrated. They work with your CFO to understand which financial levers truly matter.</p><p>The goal is to build a ranked shortlist of high-impact AI initiatives. The team is not looking for clever algorithms. They are looking for business solutions. Could AI optimize pricing in real time and potentially increase margins? Could an intelligent agent reduce the time your sales team spends on administrative tasks, freeing them up for more valuable client interactions? Could a predictive maintenance model prevent costly machine breakdowns?</p><p>Each initiative is rigorously evaluated against defined success criteria. The team looks at potential return on investment, cost of inaction, data readiness, and critically, organizational readiness. Is your team prepared to adopt a new tool? What is the training burden? By the end of these two weeks, you have a clear roadmap. You know exactly where you are going to start and why. You have defined what success looks like in concrete terms.</p><p>Phase Two: Prototype Engineering and Validation</p><p>This is where McLean Forrester separates itself from consultants who just want to sell you a report. They do not just talk about what is possible. They build it. And they do it in weeks, not quarters.</p><p>The prototyping phase is about answering a single question with empirical evidence. Is this viable? The team takes the selected initiative and builds a functional prototype. The critical difference is that they build it on your actual data. They are not using generic, sanitized public datasets to create a pretty demo. They are ingesting your real world, messy, and often imperfect data to see if the model can actually deliver value in your specific context.</p><p>This process is full of discoveries. For one logistics client who was convinced they needed a complex system to predict shipping delays, the team found that a simpler, more focused model trained on weather patterns and port congestion data was more accurate and infinitely more explainable than the elaborate solution they had originally envisioned.</p><p>The team does not just build the model and hand you a code file. They deliver a comprehensive validation report that measures both technical and business performance. They quantify the business impact in dollars and cents. They project what this prototype would mean if scaled to your entire operation. At the end of this phase, you have everything you need to make a clear go or no-go decision, grounded in evidence, not hype.</p><p>Phase Three: Production Deployment and Governance Integration</p><p>Building a prototype that works on a laptop is one thing. Deploying a secure, reliable system across an entire enterprise is a completely different challenge. Many AI projects die in the transition from successful pilot to full-scale production.</p><p>The final phase is about engineering for scale, security, and governance. The team works alongside your internal engineering, security, and compliance teams to ensure the solution is seamlessly integrated into your existing systems. In today’s environment, this is non-negotiable. Regulatory scrutiny is high, and consumer trust is paramount. You cannot afford a rogue AI.</p><p>The team puts operational controls in place. They build the monitoring dashboards that will tell you how the model is performing in the wild. They establish feedback loops so the system can learn and improve over time. Most importantly, they do not just hand you a solution and disappear. They transfer the knowledge through rigorous knowledge transfer sessions so your teams can own, maintain, and evolve the capability.</p><p>The Learning Path: Building Internal Capability</p><p>Beyond the AI Value Path, organizations serious about AI adoption need to build internal capability. The <a href="https://mcleanforrester.com/ai-learning-path">AI Learning Path</a> from McLean Forrester is designed to take business principals from basic AI literacy to an executable, AI-enabled strategy.</p><p>The program is delivered across three tiers. Foundations covers core concepts and gets you hands-on with prompt engineering. Application moves from literacy to action, helping you build an actual AI-powered workflow during the session itself. Strategy walks you through a modern strategy framework, helping you build a portfolio of ranked AI projects with value propositions and ROI analysis.</p><p>The businesses winning with AI are not necessarily the biggest ones. They are the ones led by people who decided to get capable instead of staying curious.</p><p>The Bottom Line</p><p>In a world of speculative investments and open-ended strategy retainers, a disciplined path is essential. McLean Forrester works in time-boxed sprints. They build the prototype so you do not have to guess. And they give you the evidence to make a confident decision, whether that decision is to move full steam ahead, to pivot, or to pause.</p><p>For CEOs and executive teams, the hardest part is not deciding what to do. It is deciding when to start. The technology is ready. Your data is waiting. And your competitors are already moving. The question is no longer if you should act, but how you will ensure your actions are disciplined, effective, and measurable.</p><p>Sometimes, the most valuable thing we can do is help a client stop chasing every direction at once and focus on a clear vector forward. That is the difference between burning fuel and making progress. That is the difference between activity and achievement. And that is exactly what <a href="https://mcleanforrester.com/">McLean Forrester</a> delivers for every client, every engagement, every time.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c16519d6dff6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The AI Imperative: A Pragmatic Framework for Getting Started]]></title>
            <link>https://medium.com/@michael.solidum/the-ai-imperative-a-pragmatic-framework-for-getting-started-219fa5c1c98e?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/219fa5c1c98e</guid>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Wed, 01 Jul 2026 15:54:22 GMT</pubDate>
            <atom:updated>2026-07-01T15:54:22.289Z</atom:updated>
            <content:encoded><![CDATA[<p>As <a href="https://mcleanforrester.com/about/">Heather McLean</a> of Forrester aptly notes, this question is the modern rallying cry of the small business owner. It signals a crucial awareness: AI is not a futuristic novelty, but a present-day competitive lever. However, this recognition often collides with the paralyzing fear of misstep, especially when resources are scarce and the margin for error is thin.</p><p>The answer, counterintuitively, is not to dive headfirst into strategy. Strategic planning, while vital, can become a trap, a quagmire of analysis that stalls momentum before it begins. The intellectual starting point for any <a href="https://mcleanforrester.com/where-do-i-start-with-ai/"><strong>AI strategy for small business</strong> </a>is not a grand blueprint, but a clear-eyed assessment of your current reality. For a deeper understanding of this foundational step, you can explore the <a href="https://mcleanforrester.com/services/ai-value-path/"><strong>AI Value Path</strong> </a>framework offered by Heather’s firm.</p><h3>Step 1: Diagnose Your Organizational AI Maturity</h3><p>Before you can chart a course, you must know your coordinates. Heather introduces a powerful diagnostic tool: the AI Maturity Model. This is not a vanity metric, but a strategic compass. It allows you to benchmark your organization against a clear spectrum:</p><ul><li><strong>Level 0 — The Analog Holdout:</strong> No AI use is present. Operations are entirely manual or rely on traditional software without any AI components.</li><li><strong>Level 1 — The Ad-Hoc Experimenter:</strong> AI use is inconsistent and driven by individuals. Employees might use a public tool like ChatGPT for brainstorming or drafting emails, but there is no formal policy or coordination.</li><li><strong>Level 2 — The Structured Adopter:</strong> Your business has invested in enterprise licenses for commercial large language models (LLMs). More importantly, you have established governance, training, and usage policies.</li><li><strong>Level 3 — The Data Integrator:</strong> You have deployed at least one production AI system that is integrated with your proprietary business data. This is where AI starts creating defensible value.</li><li><strong>Level 4 — The Process Re-Engineer:</strong> You have multiple production AI systems, and their adoption has led to a fundamental redesign of business processes, not just a digital facelift.</li><li><strong>Level 5 — The AI-Native Differentiator:</strong> AI is not an add-on; it is a core component of your value proposition. Domain-specific AI models are used to create a distinct competitive advantage.</li></ul><p>This diagnosis accomplishes two things. First, it demystifies the path forward by providing a tangible goal for the next step. Second, it instills a sense of progress. Moving from Level 0 to Level 1 is a significant and defensible victory.</p><h3>Step 2: Shift from “Where” to “Why” — Identify Pain Points</h3><p>Once you know where you are (Maturity Level), the question transforms from “Where do I start?” to “What is hurting me most?”</p><p>Strategy, in this context, is about deliberate focus. Avoid the temptation to look for generic “use cases.” Instead, conduct a systematic audit of your daily operations. Ask your team and yourself:</p><ul><li>What recurring tasks consume an inordinate amount of time?</li><li>Where do bottlenecks occur in our workflow?</li><li>What manual processes are most prone to human error?</li><li>Where are we burning resources on low-value, repetitive work?</li></ul><p>These <a href="https://mcleanforrester.com/catalysts-for-organizational-change-pain-points-opportunities-and-the-transformative-role-of-technology/"><strong>pain points</strong></a> are your most fertile ground for AI adoption. They represent immediate, tangible problems that AI is uniquely equipped to solve. By starting with a problem, rather than a technology, you ensure that your investment directly drives efficiency and reduces friction. To further refine your approach to identifying these opportunities, you can review the structured guidance available on the main site.</p><h3>Step 3: Evaluate Opportunities Through a Dual Lens of Effort and Value</h3><p>With a clear list of pain points, you now have a roster of potential “opportunities” to apply AI. The final step is prioritization.</p><p>This is where intellectual rigor meets practical execution. For each use case, you must conduct a rapid, honest assessment against two critical dimensions:</p><ol><li><strong>Level of Effort:</strong> What is the technical, financial, and personnel cost to implement this solution? Does it require simple prompts (Level 1 effort), or does it require integration with a CRM and database (Level 3 effort)?</li><li><strong>Value of Outcome:</strong> If successful, what is the quantifiable impact? Will it save X hours per week, reduce errors, increase revenue, or improve customer satisfaction?</li></ol><p>The aim is to identify the “quick wins,” opportunities that sit in the <strong>high-value, low-effort</strong> quadrant. These are the projects that will build momentum, generate buy-in, and fund the next phase of your journey.</p><h3>Conclusion: The First Step is the One You’re On</h3><p>As Heather aptly puts it, the starting point is not a destination, but a decision. It is the act of moving from passive observation to active assessment. By first understanding your organization’s AI maturity, then focusing on your most acute pain points, and finally prioritizing based on a clear effort/value analysis, you replace analysis paralysis with deliberate action.</p><p>This is the beginning of a true <a href="https://mcleanforrester.com/where-do-i-start-with-ai/"><strong>AI strategy for small business</strong></a>, one not built on hype, but on a grounded, rigorous, and pragmatic path to business transformation.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=219fa5c1c98e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Army AI and Autonomous Robot Boats: Redefining Pacific Logistics Through the AI Value Path]]></title>
            <link>https://medium.com/@michael.solidum/army-ai-and-autonomous-robot-boats-redefining-pacific-logistics-through-the-ai-value-path-63d45e2e6673?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/63d45e2e6673</guid>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Tue, 30 Jun 2026 15:48:30 GMT</pubDate>
            <atom:updated>2026-06-30T15:48:30.173Z</atom:updated>
            <content:encoded><![CDATA[<p>The vastness of the Pacific Ocean has always been the U.S. military’s greatest logistical headache and its most significant strategic vulnerability. Distances are measured in thousands of nautical miles, not hundreds, and the “tyranny of distance” has historically constrained operational tempo. That is why the recent statements from Maj. Gen. Gavin Gardner regarding the 8th Theater Sustainment Command’s use of Artificial Intelligence and autonomous watercraft represent more than just a technological upgrade; they signify a fundamental philosophical shift in how the Army plans to fight and sustain itself in the 21st century.</p><p>Gardner’s assertion that “if you can work in the Pacific, you can work anywhere in the world” is not hyperbole; it is a recognition that the Pacific is the ultimate proving ground for modern military logistics. However, the true brilliance of this approach lies in the realization that the private sector is already pioneering the solutions the military desperately needs. By explicitly stating he is leveraging commercial partners for warehouse management and supply chain timing, Gardner acknowledges a crucial truth: military logistics is fundamentally a business problem, albeit one with life-or-death stakes.</p><h3>The Algorithm of War</h3><p>The Army’s move to use AI for demand analysis is a significant step away from the “just-in-case” logistics that have historically clogged supply chains. Gardner pointed out the reality that resources are finite and “we just can’t afford to stock everything.” This is where Artificial Intelligence becomes the decisive factor. By utilizing AI to forecast demand over “time and space,” the Army is shifting from reactive resupply to predictive sustainment.</p><p>This is where the concept of the <a href="https://mcleanforrester.com/services/ai-value-path/"><strong>AI value path</strong></a> becomes critical. In the context of military logistics, an AI value path isn’t just about processing data; it is about creating a tangible, operational advantage. It involves mapping the journey from raw intelligence and consumption data to a deliverable action that saves time and lives. The commercial world has perfected this for inventory management, but applying it to a contested theater with electromagnetic interference, weather variability, and a near-peer adversary is a different beast entirely. The AI value path here must account for variables that Wall Street algorithms never see. By applying a rigorous value path to its logistics AI, the Army can ensure that the algorithms are not just smart, but also trustworthy and resilient against adversarial manipulation.</p><h3>The Fleet of the Future is Uncrewed</h3><p>While the AI processing data in a headquarters is powerful, the physical manifestation of this modernization is the autonomous watercraft. The news that the Army is fielding vessels over 100 feet long, capable of carrying up to eight 20-foot containers, is a game-changer. The reliance on the aging Landing Craft Mechanized-8 fleet has been a known vulnerability. The new Maneuver Support Vessel (Light), while manned, offers a glimpse of the future: speed and shallow draft. However, the focus on autonomy suggests an understanding that the future battlefield is too dangerous for manned ships in the primary logistics role.</p><p>Gardner’s vision of having “between 30 and 100” medium autonomous vessels berthed across the Indo-Pacific is the kind of distributed logistics network that makes a peer adversary sweat. Such a fleet would allow the U.S. to preposition supplies and perform “rapid insertions” of assets like the HIMARS and Marine NMESIS without putting a large, valuable target (like an LPD or LSD) in harm’s way.</p><h3>The Regulatory Bottleneck</h3><p>However, the biggest hurdle identified in this push is not technological; it is legal. Gardner correctly points to the current U.S. maritime laws requiring a minimum crew size, which limits autonomous operations to pilot programs. This is a critical friction point. For the <a href="https://mcleanforrester.com/services/ai-value-path/"><strong>AI value path</strong></a> to reach its full potential in the Pacific, the output of the AI must be the ability to act swiftly. If an AI calculates that a re-supply vessel needs to arrive at a specific island chain within a narrow window, but the ship is delayed because of a law requiring a crew that doesn’t exist or a port that won’t accept an unmanned vessel, the entire value of the AI is negated. The AI value path is broken at the last mile. Gardner’s push for the Coast Guard to accept “unmanned systems enter into ports” is a necessary step to ensure that the “value” generated by the algorithm can be physically realized on the battlefield.</p><h3>The McLean Forrester Insight</h3><p>In the current AI sphere, there is a dangerous emphasis on the “cool” factor of algorithms rather than their utility. As <a href="https://mcleanforrester.com/">McLean Forrester</a> notes, the conversation around AI is often fragmented, focusing on the model rather than the mission. The Army’s approach in the Pacific is a textbook example of how to do it right: identify the mission (logistics in a vast theater), identify the friction points (distance, resources, law), and apply AI specifically to solve those problems.</p><p>The success of the 8th TSC’s endeavor will hinge on their ability to treat the entire operation as a cohesive “AI value path.” If they can successfully integrate commercial AI for demand analysis, pair it with experimental autonomous vessels, and simultaneously lobby to change the regulatory landscape, they will have created the most efficient logistics network in military history.</p><p>The Army is right to see the Pacific as the ultimate test. It is an unforgiving environment where the rules of engagement are different. But as Gardner suggests, if the Army can untangle the knot of logistics in this region using AI and robotics, it will hold a capability that is unmatched anywhere else. It is no longer about simply moving boxes; it is about moving them with the precision and speed of an algorithm and the resilience of a robot.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=63d45e2e6673" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI Revolution 2026: Claude Surpasses ChatGPT, IBM Breaks 1nm Barrier, and What It Means for Your…]]></title>
            <link>https://medium.com/@michael.solidum/ai-revolution-2026-claude-surpasses-chatgpt-ibm-breaks-1nm-barrier-and-what-it-means-for-your-e4e689bc2bd7?source=rss-39fdc6607b28------2</link>
            <guid isPermaLink="false">https://medium.com/p/e4e689bc2bd7</guid>
            <dc:creator><![CDATA[Michael Solidum]]></dc:creator>
            <pubDate>Mon, 29 Jun 2026 15:37:30 GMT</pubDate>
            <atom:updated>2026-06-29T15:37:30.953Z</atom:updated>
            <content:encoded><![CDATA[<h3>AI Revolution 2026: Claude Surpasses ChatGPT, IBM Breaks 1nm Barrier, and What It Means for Your Business</h3><p>The past seven days in artificial intelligence have delivered a cascade of announcements that feel less like incremental progress and more like a fundamental shift in how technology will integrate into every corner of our lives. From the way video is produced to the inner workings of the human brain, the progress is breathtaking and, frankly, a little disorienting.</p><p>Perhaps the most significant narrative this week is the changing of the guard at the top of the AI industry. For years, OpenAI and its flagship product, ChatGPT, have been the undisputed leaders in the public consciousness. That story is evolving. A major report from Sensor Tower revealed that while ChatGPT recently became the fastest app to reach 1 billion monthly active users, its market share has fallen below 50 percent for the first time. The real story is who is gaining ground. Anthropic’s Claude is not just closing the gap; it is becoming the preferred choice for many businesses. The report shows that Claude generates significantly more revenue per user and converts a higher percentage of its users to paid subscriptions. Another index found that Anthropic has surpassed OpenAI in business adoption, with more U.S. companies reaching for Claude first when a new project starts. This is a stunning reversal. Businesses are choosing Claude over ChatGPT, and the reason seems clear: Claude is viewed as more reliable, trustworthy, and better suited for the demands of enterprise production work.</p><p>This trend points to a broader maturation of the AI market. The era of experimentation and novelty is giving way to a focus on practical, measurable results. This is precisely the environment where a company like <a href="https://www.blogger.com/blog/post/edit/36927854850194242/6363420211925117178#">McLean Forrester</a> thrives.</p><p>We have built our firm on the principle that AI must be implemented securely, simply, and with an unwavering focus on real business outcomes. The news this week confirms our core philosophy: the winners in this new era will not be the ones with the flashiest demos, but the ones who can successfully integrate AI into their operations to deliver tangible value.</p><p>This is where our experienced team excels. With over 30 years of enterprise technology experience, we understand the complexities of modern organizations. We cut through the hype and focus on what works, turning the overwhelming potential of AI into a clear, actionable plan for our clients.</p><p>The week’s news also saw a significant development in the field of video creation, representing a complete transformation of the production process. This aligns perfectly with our focus on collaborative innovation. We believe that the best solutions are born from close partnership with our clients. By working together, we help them identify how these new generative capabilities can be harnessed to create new value and streamline their content creation workflows, moving beyond theoretical possibilities to practical applications.</p><p>At the same time, groundbreaking research is allowing scientists to see the human brain like never before. Meanwhile, IBM has crossed the sub-1nm chip barrier, a monumental achievement that will underpin the next generation of even more powerful and efficient AI systems. This highlights how the entire ecosystem is advancing in parallel. The software is becoming more capable, the hardware is becoming more powerful, and the science is deepening our understanding of intelligence itself. However, this incredible power also introduces new risks. A bombshell report finding that AI chatbots show a left-wing bias is a stark reminder that these systems are not neutral. They reflect the data they are trained on and the choices made by their developers. As these tools become more influential, understanding and mitigating such biases is not just a technical challenge but an ethical imperative.</p><p>Japan has also firmly entered the AI race, releasing a model that performs on par with leading systems. This is a crucial development, diversifying the global AI landscape and ensuring that the technology is shaped by a wider range of perspectives and values. As AI becomes a tool of national and economic strategy, this kind of healthy competition is vital. On the other end of the spectrum, China’s delivery robots are becoming a part of daily life, offering a glimpse into a future where intelligent machines handle logistics and mundane tasks, allowing humans to focus on higher-level work.</p><p>It is this principle that guides everything we do at <a href="https://mcleanforrester.com/">McLean Forrester</a>. Our approach is grounded in insight and real-world experience. Our team brings 60 years of combined military and government experience, which instills a “mission first” mindset and a deep understanding of security, risk, and operational efficiency. We view AI adoption as a strategic logistics problem, ensuring that we automate the right processes in the right way to deliver a decisive advantage.</p><p>As we process these developments, it is also wise to keep a critical perspective. The reported left-wing bias in chatbots is a powerful reminder that these systems are not neutral arbiters of truth. They are products built by people with inherent, if unconscious, biases. As leaders and decision-makers, we must be aware of these tendencies to ensure the technology is used responsibly and equitably. The news of Claude’s rise over ChatGPT also underscores the need for a multi-model strategy. The market is not a zero-sum game. Many enterprises are running both models simultaneously to leverage their respective strengths. This speaks to the need for an agile and informed approach to AI adoption, one that a accessible and direct partnership with a company like ours can provide.</p><p>ChatGPT’s upcoming voice mode sounds truly incredible, promising to make interactions with AI even more seamless and natural. As interfaces evolve, the ability to adapt and integrate the best tools for the specific job will be crucial. We are built to help our clients navigate this complex and fast-moving landscape, providing them with the clarity and speed needed to stay ahead.</p><p>Built on Vision. Driven by Innovation. McLean Forrester started with a simple conversation in a kitchen. A group of experienced professionals came together with a shared vision to build something extraordinary beyond the limits of corporate and government structures. United by expertise and a passion for innovation, we set out to create a technology company focused on people and real customer value. To learn more about our team and our approach, visit our <a href="https://mcleanforrester.com/about/">About</a> page and discover how we are making AI work for businesses.</p><p>This was another major week in AI. The progress is undeniable, but it is the application of this technology to solve real-world problems that will define its ultimate value. That is the gap McLean Forrester is built to bridge. We provide the strategic edge, the technical expertise, and the commitment to results that allows our clients to turn the promise of AI into tangible performance. Visit our <a href="https://mcleanforrester.com/">homepage</a> to see how we can help you navigate the future with confidence.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e4e689bc2bd7" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>